Hypoxia is a major stress on human physiology, which elicits an acclimatization response consisting in an increase in hemoglobin (Hb) concentration, which in turn compensates for the decrease in oxygen saturation of Hb (O2 sat). Paradoxically, Tibetan highlanders present markedly low O2 sat, but relatively little increase in Hb levels resulting in a dramatic reduction of arterial oxygen content. Recent studies have identified advantageous alleles that are consistently associated with lower Hb levels, suggesting that natural selection in Tibet favored variants that counteract the acclimatization response. Therefore, despite these advances, how Tibetan physiology copes with such low arterial oxygen content remains entirely unknown. In preliminary studies, we have shown that the Tibetan gene pool results from the admixture of ancestral populations with different degrees of adaptation to hypoxia and that selection acted since admixture to increase the frequency of hypoxia-adapted alleles. This setting allows a set of powerful approaches that look for genomic regions with excess ancestry from the adapted population. We will apply these and other standard mapping approaches to a new collection of >1,000 DNA samples from Tibetan women collected in Nepal. Importantly, extensive data are available for hypoxia-response traits and reproductive variables. In turn, the alleles identified through these analyses will be investigated in detail to learn abou important parameters of the evolutionary process. To this end, we propose the following specific aims: Aim 1. We will genotype ~900,000 SNPs/CNVs in 1,016 Tibetan women and 100 newly collected Sherpa samples. We will infer local ancestry in the Tibetans to identify regions with excess hypoxia-adapted ancestry. Aim 2. We will test the long-standing proposal that variation in Hb levels and O2 sat is adaptive by testing the correlation between these hypoxia-response phenotypes and reproductive success. In addition, we will use the same genotype data collected in Aim 1 to map Hb levels and O2 sat by association and admixture mapping. Aim 3. We will use the same genotype data collected in Aim 1 to map variants contributing to variation in reproductive fitness. We will also use the reproductive data to provide direct estimates selection coefficients for the variants discovered in Aims 1 and 2. Aim 4. We will sequence the genome of 6 Sherpa and 6 Dai and will use these data to model the demographic history of Tibetans and of the ancestral populations contributing to their gene pool. We will then test for signatures of natural selection and use the inferred demographic model to estimate selection coefficients for different sets of variants of interest. Through the proposed work, we expect to gain new insights into the biology and genetics of hypoxia response. In addition, the reproductive data will allow us to elucidate the adaptation process with unprecedented resolution, thus making Tibetan adaptations to hypoxia a new thoroughly understood case of natural selection.